future power grids able to direct when fridges go into their "off" cycle early?

Yes, we do have the technology to send activation messages through the power grid. . . and it is widely used in Europe, mainly to change the setting on the electric meters after hours, etc . . . . To install the required sensors on washing machines, air conditioners, etc. is not sheep. . . . and who will monitor the users who could easily by pass those devises??? and the hospitals, older peoples, etc??? You will have laws that would be a nightmare to enforce . . . .

Also, usually the overloads on the power grid is coming more from industrial usage then residential . . . . unless it is super hot weather, etc . . .

In Europe, the activation device in the meter is an integral part of it and can not be tempered with . . . .

This already exists at least experimentally, and possibly in service by now. The signal which is used is the power line frequency, which drops when the grid is under a load stress, and raises when there's an excess of generating capacity over load. For example, on our 50Hz system, below 49.9Hz, fridges could increase their running temperature by a degree or so, and below 49.8Hz, they could decide not to operate at all.

For something which can store energy for later use or to avoid later use, it can monitor for frequencies above the nominal and draw extra power at those times. An example might be a stored water heater which would heat up a larger volume of water if the frequency is above 50.1Hz. A freezer might reduce its set temperature by a few degrees so it can later reduce its consumption when the frequency drops back to nominal or drops below nominal.

There's a lot more to it than this, e.g. if all fridges switch off and on at the same time, that would probably cause much more of a problem than not having the system at all, but this sort of thing can be tackled by introducing randomness and damping into the equation.

| Yes, we do have the technology to send activation messages through | the power grid. . . and it is widely used in Europe, mainly to change the | setting on the electric meters after hours, etc . . . . To install the | required | sensors on washing machines, air conditioners, etc. is not sheep. . . . | and who will monitor the users who could easily by pass those devises??? | and the hospitals, older peoples, etc??? You will have laws that would be | a nightmare to enforce . . . . | | Also, usually the overloads on the power grid is coming more from industrial | usage then residential . . . . unless it is super hot weather, etc . . .

Most new A/C systems are under electronic control these days, with digital temperature settings. Under grid signal control, they could, in areas seeing power shortages, incrementally increase the effective setting. So if you set it to 76 F, under shortage conditions, a signal could cause the effective setting to be 4 degrees higher, thus 80 F.

There needs to be legal ways to bypass this. And enforcing against illegal bypasses is not going to be easy. One example is a home with someone who has a health condition that requires a temperature ceiling.

What I think is the best answer to this is to migrate to time based pricing of usage. There would be a base price level. Utilities would then send a signal out (relayed via the meter) indicating the pricing adjustment based on supply/demand. This would be a percentage signal. Then appliances can receive the power price adjustment level and make decisions to shut off or not based on owner/operator preference.

Higher pricing should be the means of encouragement to participate. An A/C unit could be set to a factor for how many percents of price increase will adjust the temperature one degree. Someone who wants to save more can set it to 1 degree per percentage point. Others might set it to 1 degree per 4 percentage points. Someone with a health issue might set it to 0 degrees adjustment. Financial assistance programs can be used to deal with cases of low/no income people who also have health issues by supplementing their finances or arranging special pricing through the utility.

Digital power meters would be addressable devices and report usage by time back to the utility, so correct calculations can be used. Signalling SHOULD be UNencrypted so it can be monitored for correctness, but be verified by means of an encrypted hash code to verify the signal received is truly from the power meter or from the utility. Encryption should also be an option in cases where the customer wants to ensure secrecy of their business relation with the utility. The power meter would then provide the utilization signal on the load side by means of a signal on the power lines that can also be tapped off by a filter and utilized in other ways.

| In Europe, the activation device in the meter is an integral part of it and | can | not be tempered with . . . .

You hope.

But I think the economic incentive method would be more appropriate. It removes the need to tamper for most aspects. Sure, someone might want to tamper to make their day usage appear as night usage. But with meters that can be polled at any time and will upload the usage data right up to that minute, it would not be easy to fool it even if it can be tampered. Of course the devices would be well sealed and the signals unspoofable even though they should be monitorable.

There are too many cases where shutting things off too much can cause more problems that I would not go so far as to have blanket requirements. For example refrigerators might have to run anyway, but just run less, to keep the temperature under a certain requirement. A lot of other practices like opening the door less often and filling empty spaces with water bottles is going to do a lot to avoid power usage.

On 17 Jan 2007 21:29:38 GMT Andrew Gabriel wrote: | In article , | "Jevan Pipitone" writes: |> |> This is an idea I came across a few years ago. With the recent melbourne |> blackout and the issue of suddenly needing power but not being able to |> fulfil it from backup generators etc, this could be another possible |> contributor. |> |> "Some" Electrical devices could have built in technology that enables them |> to be shut down from the power grid. That is, the electrical authorities if |> there is too much power being consumed, could send a message down the lines |> (such as by changing the phase slightly of the power, or something) which is |> detected by these electrical devices, which then temporarily shut down. For |> example, the fridge could go into its "off" cycle earlier. | | This already exists at least experimentally, and possibly in service | by now. The signal which is used is the power line frequency, which | drops when the grid is under a load stress, and raises when there's | an excess of generating capacity over load. For example, on our 50Hz | system, below 49.9Hz, fridges could increase their running temperature | by a degree or so, and below 49.8Hz, they could decide not to operate | at all. | | For something which can store energy for later use or to avoid | later use, it can monitor for frequencies above the nominal and | draw extra power at those times. An example might be a stored | water heater which would heat up a larger volume of water if the | frequency is above 50.1Hz. A freezer might reduce its set temperature | by a few degrees so it can later reduce its consumption when the | frequency drops back to nominal or drops below nominal.

Those seem like rather gross frequency shifts. I guess it depends on what kind of demand load is being referenced. One issue with this is that it also limits the sharing of power w/o a frequency converter. For example, should a refrigerator in Norway be mitigated when there is a heat wave in Italy?

I'd recommend a pricing signal and time frame metering with a resolution at least as fine as one hour, if not finer. The signals would indicate the percentage above or below the nominal pricing with a resolution of at least one percent, if not finer. Then let devices adjust their usage to the pricing based on owner/operator preference. Utilities would have to be required to average the signals over some period of time such that off peak utilization brings the long term average to nominal. That can't be done per day, necessarily, as in some locations, demand can even be high at night (walk through a Houston Texas neighborhood at 3 AM on a hot July night and hear the drone of nearly every A/C system running).

Time based electricity pricing for everyone, combined with social programs to assist low/no income people where needed, would be the way to go, IMHO.

BTW, this will also need to be added to computer PSUs to provide the CPU with a port to read the electric pricing signal right off the mains line. Computer can be programed to shutdown, if desired, at certain price levels. Or they can be in control of other devices. The signals, and the port interface, need to be open unencumbered non-proprietary standards, with sufficient verification security (digitally signed SHA256) so as to be unspoofable.

No!!! we can not play with the frequency!!! the frequency of the grid MUST be as constant as possible. Usually, the largest power plant on the grid will be used to solidly maintain the frequency, and I believe every day or two they compensate for the accumulated lead or lag. The others smaller plants on the grid must follow that frequency or find their generators having large swings of power in or out . . . . . a small fraction in the frequency represents a "slip" in the synchronous generator . . . . .

Some utilities have offered a tariff for this sort of thing for a while now. I know Detroit Edison offered that for electric hot-water heaters and air-conditioners for a while. It allowed the utility operator to 'shut-off' the large loads during times of peak demand. System operators had a prediction of how much load it would drop based on the outside temperature (for the A/C). Mild days it wouldn't save much, but really hot days it could be several MW of peaking power equivalent.

But some folks have cheated the tariff by rigging bypass circuitry. So they would get the rate relief from the utility, and never have their power interrupted. Those that were caught had their back billing re-calculated as if no such rate relief existed. Stood up in courts, they paid.

As data communication technology over power lines improves, this sort of thing could get better and better.

daestrom

For 30 or 40 years, maybe more, we in the south of the world have had timed hot water heaters in some places with of course reduced tarifs and other areas have had Zellweger relays which allow the sys.op. to turn groups of loads on and off when he needs them. Again this is only water heaters and street lighting loads that he can arbitraily switch with out too much trouble. The switching is controlled by a 1050hz tone (think that is number) with I believe 25 channels controlled by timing between 1050 tone bursts on the line.

-- John G

You're missing the point. No one is deliberately playing with the frequency. It changes naturally anyway depending if there's an excess or deficit of supply over load. It thus provides a direct indication to appliances of the status of energy supply, and in the case of gross mismatches, it gives enough advanced warning of load shedding that the event could possibly be significantly mitigated.

I have give an example of significant frequency shift due to supply/load mismatches in the UK in the following article:

----------

Actually, except in a few small systems, the largest unit on the grid does not regulate frequency. If it did, it would have to carry all variations in load and this is, for a large system, both impractical and uneconomic. This is a practice which, in most places, is long gone. Load frequency control in which normal governor settings determine the initial load sharing and the change in frequency with load is followed by adjustment to correct frequency and adjust load to the desired sharing.

It is true that by the time that you can detect a frequency difference between machines, you are in deep doodoo but that is not an issue that is a problem with normal load changes.

By the way time of day metering as indicated by phil --nonews is in fairly common use. No one is forced to shift load but the economics of doing so are apparent. Control through the pocketbook. I also believe that about 20-30 years ago there were some setups where selected domestic loads were on a separate circuit which could be interrupted at peak periods. Probably pocketbook control was cheaper to implement and just as effective.

Don Kelly snipped-for-privacy@shawcross.ca remove the X to answer

----------------------------