How is the US power grid kept in phase?

[stedwards]

A quick calculation shows that across the continental United States, spanning about 3000 miles, there is a delay from east to west coast of about 1/60th of a second. 3000 miles divided by 186,000 miles per second. This is a full cycle phase variation cost to cost.

Is there truly "The Power Grid", as it is so commonly referenced, interconnected across the US, or is it divided into sections that can be kept more-or-less in phase?

 

[Simon Bridge]

There is no need for the entire network to be in phase.

But you are correct in thinking that "the power grid" is an over-simplification.
Power networks are divided up into sub-networks and so on just like computer networks.
Calling an entire continental power network "the grid" is like calling all internet connected computers "the cloud".

 

[Jeff Rosenbury]

Keeping the grid in phase is not an easy task. In addition to the propagation phase shift you mentioned, there are other phase shifts.

The big one is the transmission phase shift. AC transmission lines don't use a higher voltage to shift power "downhill" like a simple DC model might suggest. Instead, the amount of power moved depends on the phase difference between one side of the transmission line and the other.

To some extent this is self regulating since a lagging area starts to draw more power until it catches up. On a linear line, that would work fine. But we have a roughly two dimensional grid. Thus there's the potential for power to start chasing itself around in circles. Power companies have control centers and equipment to stop this and other problems.

 

[jim hardy]

Relativity aside,

synchronous generators that are connected to one another rotate in "lock step" . Their rotors remain within a fraction of a turn of one another. One that tries to pull ahead or behind the others will give or take power through its wires as required to stay in step. That's just the way Mother Nature made synchronous machines.

Regions of the US and Canada adjust their steam valves* to keep energy flowing between them and their neighbors where they want it.
*(water valves for Hydroelectric plants)
http://www.eia.gov/todayinenergy/detail.cfm?id=4270

The map above below_jh shows that electricity tends to flow south in North America. The numbers on the map reflect average net power flows—metered hourly—between electric systems aggregated by regions for the year 2010. Most electric power demand is served by local generators. Net interregional trade accounted for less than 1% of delivered power in 2010. However, excess, low-cost power—primarily from hydroelectric generators in the Pacific Northwest, Manitoba, and Quebec—supplied higher-cost markets to the south.

In the map above, the numbers next to the arrows represent annual net flows of electricity between regions measured in millions of megawatt-hours.

 

[jim hardy]

You can watch the angular displacement here

http://fnetpublic.utk.edu/anglecontour.html

 

[Jeff Rosenbury]

You can watch the angular displacement here

http://fnetpublic.utk.edu/anglecontour.html

What a nice map.

I noticed the Dakotas have a lot of variation. I assume that's due to wind power.

 

[the_emi_guy]

I agree with your suspicion about the frequency offset chart.

It has been showing the western third of the U.S. running 0.020Hz fast for the past 5 minutes. This would add up to 6 entire cycles of phase, yet the phase angle chart never moves beyond 30 degrees in this area.

 

[stedwards]

I can see how two generators, running out of phase would act like a motor-generator pair. Also, the US power grid seems interconnected coast to coast, if we are to believe the charts.

However, for two oscillators, time delayed by distance, I can't quit see it. Each will see a signal that is behind in phase, and so try to pump energy to the other. This seems to be a contradiction---which is dumping energy to which?

 

[Windadct]

The FNET site is interesting --at this point I see the middle of the country being the most stable, and the other regions oscillating - it would seem energy "sloshing around" a little.

As for the 5 minutes at -0.02 perhaps some issues with sampling or data collection. Also with the data supposedly at the 120V Receptacle - that is really quite removed from the transmission system, so how such large regions could really be matched that well is beyond me...

 

[anorlunda]

Jim Hardy gave the key fact. Power sent from a generator to the grid is proportional to the sin of its angle.

First, consider the consequences of that. To have a steady state, power not changing, then the rate of change of the sin of angle is zero, thus the rate of change of angle is zero. That expresses the conditions for a phase lock. In order for power flows to be constant, all generators in the interconnection must run phase locked. That defines a synchronous grid.

Note that the slope of a sin goes to zero, then negative as the angle approaches 90 degrees. That is called transient instability. If that happens to one generator, it looses synchronism and becomes unstable. Protective relays will trip it to remove it from the grid.

Next question, angle relative to what? In a grid with hundreds of thousands of nodes, there is no simple answer. An approximate answer is relative to electrically nearby neighbors. The actual calculation is called load flow, or power flow. In a load flow calculation, one aribtrary node is assigned the angle zero. I used to earn my living writing load flow and dynamic transient simulation software.10 years ago, there were three synchronous grids in the U.S. and Canada.. The eastern interconnect (roughly U.S. and Canada east of the Rockies), the western interconnect (roughly U.S. and Canada west of the Rockies) and Texas. They were connected to each other only by D.C. Lines which are not synchronous. Texas did not want federal regulation, so they abstained from interstate transport of power. Things may have changed in the past 10 years.

In principle, the bigger the grid the better. In case of emergencies, my neighbors can support my part of the grid. Then the neighbors have neighbors and so on. Small isolated grids, such as on an island, have more trouble maintaining voltage and frequency and reliability.

 

[jim hardy]

This seems to be a contradiction---which is dumping energy to which?

I'm hardy a resource for relativistic effects.
Are not things that interact at speed of light considered simultaneous ?

So we get into observer frames .

At c, Miami is ~10 milliseconds away from Ft Collins Co . So we added 10 milliseconds to WWV time as received on our radio and set our computer clock to that.

An observer halfway between us, somewhere around Little Rock, would see our clock in synch with WWV. And he'd see our generators and those in Ft Collins ahead and behind him by equal amounts.in synch too. Observers anywhere else would see our clocks out of synch, and our generators having unequal phase differences wrt himself.

But the kilowatts flowing along a hypothetical lossless power line from Miami to Ft Collins would have to be constant - ie all observers along it would see volts and amps having same phase relation to one another , even though they'd see different angles with respect to the ends of the line.
So they'd all see power flowing in the same direction., regardless what their local clocks report for the time of day.

So - is there such any thing as a universal time ? Did Einstein ponder the grid?

As stated, I'm hardly a resource for relativity.

 

[Jeff Rosenbury]

I can see how two generators, running out of phase would act like a motor-generator pair. Also, the US power grid seems interconnected coast to coast, if we are to believe the charts.

However, for two oscillators, time delayed by distance, I can't quit see it. Each will see a signal that is behind in phase, and so try to pump energy to the other. This seems to be a contradiction---which is dumping energy to which?

The motor/generator does not see the other motor/generator. It sees the end of the transmission line. So it doesn't see what's happening to the other motor generator, it sees what happened a fraction of a second ago.

The transmission line holds a lot of power which will buffer any sudden changes. So if both are "in sync" (whatever that is) the line will just float. If one starts to lag, the line will pay the power price until it can pull the power from the other motor/generator to replenish.

Nor is the transmission line a single thing. It's energy is distributed in the distributed inductance and capacitance of the line, acting somewhat like a long battery.

So if a motor/generator is acting like a generator, then the local mechanical power (the prime mover) is cut, there will be some slip in the phase while the line adjusts. This would be a serious problem if the line were actually 2,000 miles long, but the grid is a grid, so no transmission line is actually that long. They like to keep the phase angle of a single line quite a bit less than 90º as I recall from my power class some decades ago.

I do recall being told of warped generator shafts from sudden phase shifts, but I think those might have just been war stories. Steel is pretty tough.

 

[jim hardy]

I do recall being told of warped generator shafts from sudden phase shifts, but I think those might have just been war stories. Steel is pretty tough.

There's protective relays to prevent that.
We did have one generator that failed to disconnect in a severe underfrequency event and local grid collapse. Its turbine shaft sheared. A friend made a workshop out of the LP turbine upper casing.

 

[Finny]

Boaters plugging into marina electricity dock outlets are cautioned not to use a "Y" power cord to connect into two 120 volt outlets to form 240 volts for the boat. I don't remember where I came across that folk 'lore'.

These ' Y 'connectors can be used 'backwards' combine two 120 volt leg's to form 240 volts required for some systems, perhaps air conditioning, for example. Their intended use is to provide 2 x 120 volt outlets from a 240 volt source, to boats only wired for 120 volts.
The reason usually given is that two such 120 volt sources might be 'out of phase' and lead to unwanted voltage variations, especially if the two 120 volt legs are from different higher voltage on shore lines. That never made a great deal of sense to me, and seeing how closely the frequencies are across the county it makes less sense now. Absent catastrophic conditions, are local frequency variations ever far enough off to cause issues? Seems absurdly unlikely.

 

[anorlunda]

Boaters plugging into marina electricity dock outlets are cautioned not to use a "Y" power cord to connect into two 120 volt outlets to form 240 volts for the boat. I don't remember where I came across that folk 'lore'.

These ' Y 'connectors can be used 'backwards' combine two 120 volt leg's to form 240 volts required for some systems, perhaps air conditioning, for example. Their intended use is to provide 2 x 120 volt outlets from a 240 volt source, to boats only wired for 120 volts.
The reason usually given is that two such 120 volt sources might be 'out of phase' and lead to unwanted voltage variations, especially if the two 120 volt legs are from different higher voltage on shore lines. That never made a great deal of sense to me, and seeing how closely the frequencies are across the county it makes less sense now. Absent catastrophic conditions, are local frequency variations ever far enough off to cause issues? Seems absurdly unlikely.

Quite true, but a different meaning of the work phase than we are discussing here.

 

[Finny]

"Quite true, but a different meaning of the work phase than we are discussing here."

Which part...absurdly unlikely or that it is a real problem? Thanks

 

[anorlunda]

"Quite true, but a different meaning of the work phase than we are discussing here."

Which part...absurdly unlikely or that it is a real problem? Thanks

i

In a three phase power system. There are three wires. Each "phase" carries voltage 120 degrees different from its neighboring phase. But consider a three phase generator sending power via a transmission line to a load. In that case all three phases at the generator end might lead the corresponding voltage at the remote end by one degree. We call that a phase lead of one degree.

The marina was warning about wiring errors, which of the three phase wires are connected to what. In this discussion we ar etalking about the angular difference (phase) between voltages at one end of a wire from voltages at the other end of the wire. Those are two different uses of the word phase.

I hope that clarifies. Explaining things verbally without use of a blackboard to draw pictures can be difficult.

 

[Finny]

anorlunda posts:
"The marina was warning about wiring errors, which of the three phase wires are connected to what. In this discussion we ar etalking about the angular difference (phase) between voltages at one end of a wire from voltages at the other end of the wire. Those are two different uses of the word phase..."

Ah, I understand. I was not clear in my post. I was intending to address single phase systems. And any possible phase differences exactly as you describe it...and I think I answered my own question thanks to your clarification.

The three phase systems you mention would clearly be a problem with phase issues and in fact their delivery of 208volts instead of single phase 240 volts is also a problem. Boasters can tell if they are connected to a three phase system by checking their voltage monitoring aboard. Many will not use such a low voltage as it can cause motor problems.

What I was wondering, if there was any realistic problem of different phases due to,say, the possibility of two
legs in a pedestal being supplied from different sources, say one local, and the other distant, say from Canada. That's a better description of the scenario I was proposing and why it seems "absurdly unlikely to me".