What's the continent-wide frequency, phase angle and voltage, Kenneth?

April 10, 2011 4:34 AM

FNET is a low-cost, GPS-synchronized wide-area power system frequency measurement network. Highly accurate Frequency Disturbance Recorders (FDRs) [...] measure the frequency, phase angle, and voltage of the power signal found at ordinary 120-V electrical outlets. Animated map. [via]
posted by slater (26 comments total)

As a guy who used to work on standards for synchronizing cellular networks using GPS, I can say: This is really cool.

Now I'll be watching that animated map all morning waiting for my power to reach 60.000 Hz.

Thanks, slater!
posted by ZenMasterThis at 5:20 AM on April 10, 2011

The Wiki page and the Uni page seem to give different impressions; the Uni page makes no mention of why they are doing this, so you're left with a vague impression of "something neat". The "sample events" are interesting to be sure - but what do they mean? Are they tied to anything other than "here's a data glitch"?

Power line systems already have monitoring systems that keep track of power loads, distribution, and problems (line breaks, sags, overloads, etc).

For the typical home owner, a Kill-O-Watt device might be more useful (but it's not the same, but loads more practical).
posted by Old'n'Busted at 5:21 AM on April 10, 2011 [1 favorite]

With so few instruments in the network I'm having a hard time believing any of these results are meaningful. The "islanding", esp. near the southern tip of Illinois, looks much more like an artifact of network structure and the matrix interpolation method than any real topology.

Still cool to look at though. Building this network up sounds like a a job for the DOE or DHS. A shame they don't often deal in the elegant and cheap.
posted by clarknova at 5:40 AM on April 10, 2011 [1 favorite]

Building this network up sounds like a a job for the DOE or DHS. A shame they don't often deal in the elegant and cheap.

DOE or DHS will buy it when Raytheon, L3 or Smiths Industries tell 'em to. At 50x the price. They'll even throw in a nice, fat VP position in the private sector!
posted by kcds at 6:33 AM on April 10, 2011 [2 favorites]

Following the widespread blackouts on the US/Canada eastern seaboard on August 14 2003, one of the NERC recommendations [pdf] was to use time-synchronised data recorders to provide a better view of the real-time state of the transmission grid. The FNET system is just one of many systems installed in response to that recommendation. The North American SynchroPhasor Initiative has much more information.

The 'found at ordinary 120-V outlets' statement could be misinterpreted. Yes, these devices are plugged into a standard outlet, but the real value of their data is for the transmission (thousands of volts) electricity network. The elegance of the FNET solution is it's simplicity and cheapness - it only has to measure at 120V and does not have to be installed in an electricity substation and connect to very expensive high voltage measuring equipment.

Putting one of these devices in every home would not add much more value, either to the grid operator, or to the consumer. The benefits of synchrophasor measurement is to allow a macro-level view of the transmission network to allow system operators to monitor and maintain equilibrium on the network. That's why the data is so coarse (geographically) - it has no need to be any more fine-grained.

I'm really looking forward to the day when we can use and visualise the data flowing from modern 'smart' meters - then we will truly have a real-time model of the entire electricity network from generators through to 'mom and pop' consumers.
posted by KirkpatrickMac at 6:52 AM on April 10, 2011 [4 favorites]

Quebec's always going to be doing its own thing, as it has HVDC links and maintains its own frequency reference. This scheme is a bit more complex than the beautiful 1940s Synchronome that Ontario used to use as its time sync. It now hangs in the boardroom of the IESO in Clarkson.
posted by scruss at 6:56 AM on April 10, 2011

I've watched this for a few minutes now and all the areas are flashing different colours. All areas bar one.

Yes, ladies and gentlemen, Texas is slow...
posted by sodium lights the horizon at 7:11 AM on April 10, 2011

oh, very cool.
posted by rmd1023 at 7:21 AM on April 10, 2011

the beautiful 1940s Synchronome

That reminded me of a project I was once part of - to refurbish some really old (early 1900s) electricity substations for a railway company in the UK. One of the substations had a device called a 'frequency converter' - a motor, powered from the mains voltage at 50Hz, which rotated and geared down to a driveshaft at 33rpm, which in turn rotated a substantial generator that produced 240V at 33Hz.

Turns out the only thing powered from this frequency converter was the clock in the nearby railway station manager's office - the clock was reliant on 33Hz in order to maintain correct time. In past times, the railway signalling system also ran on 33Hz, but it had all been replaced some years earlier with equipment that ran directly at 50Hz. So, for some tens of years, this (large, expensive) frequency converter had been continually maintained for the sole reason of providing the correct time.

I was never happier than when I presented the station manager with his brand new, AA battery powered quartz clock and I got the frequency converter switched off.
posted by KirkpatrickMac at 7:24 AM on April 10, 2011 [13 favorites]

Geez, that map makes me nervous. Are frequency changes really rippling through the system that quickly?

It does show you how sensitive our grid really is. I heard secondhand of a network operator being able to see little blips on the load graph every morning at 5:00 (when all else was quiet) and successively smaller blips at 5:09 and 5:18. The only explanation they had was that this was load from alarm clocks going off and the successive blips were from people hitting the snooze button.

Of course this could just be a power industry joke. If so, it's the only one I know that's not dirty.
posted by RobotVoodooPower at 7:53 AM on April 10, 2011 [2 favorites]

The FBI is interested in this

It seems a small leap to imagine that the power-grid operators are hashing time information in those variables (frequency, phase angle, and voltage). It could also be that there is simply enough natural variation in the power signals that an entity like the FBI need only keep a recording of the various power grid patterns.
posted by kuatto at 8:44 AM on April 10, 2011

KirkpatrickMac's comment reminds me of a news story from a few days ago: a company called American Superconductor (AMSC) got their stock price zapped by over 40% after their largest customer, Chinese state-owned Sinovel Wind Group Co., Ltd., decided to cancel a big order and not pay for past deliveries. One of the interesting things AMSC makes with their high temperature superconducting wire (in addition to more efficient wind turbines) is a reactive power device - or giant buffer that keeps the grid stable. (ps. happy 100th birthday superconductivity!).
posted by HLD at 11:01 AM on April 10, 2011 [1 favorite]

The load blips could be from people getting up, turning on coffee heaters or drawing hot water from electrically-heated tanks… that would still produce the same pattern of blips (and is easier to believe than that the alarm clocks themselves produce a noticeable blip!)
posted by hattifattener at 11:06 AM on April 10, 2011

"Putting one of these devices in every home would not add much more value, either to the grid operator, or to the consumer. The benefits of synchrophasor measurement is to allow a macro-level view of the transmission network to allow system operators to monitor and maintain equilibrium on the network. That's why the data is so coarse (geographically) - it has no need to be any more fine-grained.

I'm really looking forward to the day when we can use and visualise the data flowing from modern 'smart' meters - then we will truly have a real-time model of the entire electricity network from generators through to 'mom and pop' consumers."

By "value", do you mean cost? Or do smart meters measure something different from what is being shown in the FNET map?

Also, for the uninitiated, like me: What comprises the separated areas: east US, west US, Texas, Quebec, delineated by the brown lines? And the representation of Hz, what does that mean? I know what Hz measures, but why use Hz to display...well, what? Does electrical power load cause frequency to fluctuate?
posted by Xoebe at 11:59 AM on April 10, 2011

If you watch just one region, you will see it oscillate back and forth between plus and minus 60.0 Hz. These short term differences are very tiny, about 5 microseconds per cycle max, but they always average out to exactly 60 Hz.

The frequency over time is very accurately regulated so that errors in frequency do not accumulate. Over a 10 minute period, one hour, one day or one year, the total number of cycles is corrected to be exactly right, 60 per elapsed second. If the system runs slightly fast for one minute, it is corrected to run slightly slower the next minute. This is why a $10 kitchen will clock with a synchronous motor running on AC will have greater accuracy than a $1000 chronometer over a year's time. The kitchen clock is constantly readjusted to be exactly right. The chronometer accumulates errors forever until reset manually.

The fluctuations are caused by momentary changes in the load on the system. The input power to the generator (steam or hydro, for example) must exactly balance the output power. When a new power load is connected to the system (by turning on lights or motors, for example), the generator starts to slow down slightly until the feedback controls can increase the input power. Likewise when a new power load is removed from the system, the generator speeds up slightly until the feedback controls again adjust. The momentary speeding up or slowing down of the generator causes the frequency fluctuations.
posted by JackFlash at 12:15 PM on April 10, 2011

As an analogy, you can think of the cruise control in your car. You set it for exactly 60 MPH and every thing is fine until you get to a slight hill. The engine isn't putting out enough power to compensate for the hill so the car slows down to 58 MPH. The cruise control sees the speed slowing and reacts by increasing the throttle to provide more power to get you back to 60 MPH. As you go over the top of the hill, there is less load so the car starts to speed up to 62 MPH. The cruise control sees the increasing speed and backs off the throttle to provide less power until the speed is 60 MPH again. It is the brief lag between the load change and the adjustment of the throttle that causes the small fluctuations in speed.

The same sort of fluctuations happen to the generator when the electrical load on the system changes. The generator must spin at a precise number of revolutions per minute to generate 60 Hz. As the speed of the generator fluctuates, the cycles per second of the electrical power fluctuates slightly from 60 Hz.
posted by JackFlash at 12:34 PM on April 10, 2011

> One of the interesting things AMSC makes with their high temperature superconducting wire

I don't think D-VARs use superconductors. Really big caps, maybe.
posted by scruss at 2:18 PM on April 10, 2011

GPS-synchronized wide-area power system frequency measurement network

Too many nouns in this string.
posted by Mental Wimp at 4:39 PM on April 10, 2011

Also, here's a neat realted concept: hysteresis.
posted by Mental Wimp at 4:42 PM on April 10, 2011

"realted" = "related". sheesh.
posted by Mental Wimp at 4:43 PM on April 10, 2011

For this to work, all the measurement devices have to sync their clocks pretty darned accurately, microseconds not milliseconds. How do they do that? Thinking about that stuff makes my head hurt.

Years ago, I worked with an electrical engineer who told me about a job where they brought a new high voltage line into a refinery from a different part of the grid straight from the Bonneville Power Administration. The existing line ran up the coast from the south. Because the two lines ran several hundred miles over very different routes, they were different lengths and had different inductances. When they were brought together, they were out of phase. I have no idea how they fixed it.

Next time you are driving along a road paralleled by a high voltage line (these are the big ones that feed substations), every so often you will see a pole or tower with a different arrangement of supports and insulators so that two of the lines will switch places. These are used somehow to tune the lines.
posted by warbaby at 5:22 PM on April 10, 2011

For this to work, all the measurement devices have to sync their clocks pretty darned accurately, microseconds not milliseconds. How do they do that? Thinking about that stuff makes my head hurt.

The FNET Frequency Disturbance Recorders contain GPS receivers. They use the timing signals from the Global Positioning System to synchronize their clocks. GPS timing accuracy is on the order of 10 nanoseconds (GPS satellites are essentially atomic clocks with transmitters).
posted by RichardP at 5:47 PM on April 10, 2011

GPS works by measuring the speed-of-iight delay between you and a handful of satellites. As a side effect, a GPS receiver gets a very precise notion of what time it is. I wouldn't be surprised if GPS is used as much for time transfer as it is for positioning.
posted by hattifattener at 7:34 PM on April 10, 2011

Is this fluctuation in the Hz the reason why the second hand on an analog clock seems to tick a hair slower or faster from time to time?

(yes, I have been EXTREMELY bored enough to notice this)
posted by roboton666 at 9:23 PM on April 10, 2011

Warbaby, I was taught that the reason high voltage wires were rotated like that was that air is not a perfect insulator, and each of the three phases has to spend equal time nearest to the earth (if the wires run in an equilateral triangle arrangement) or in the middle (if side-by-side) or the voltage delivered to the load will not be equal on all three legs due to asymmetric leakage.
posted by tigrrrlily at 10:44 PM on April 10, 2011

That reminds me, for what its worth, my coffee gains time. Actually the last two coffee makers that I've had gain time. (show it to be later than it is) I suspect they are designed for a 50 cycle grid instead of the 60 cycle system I live in here in Texas. I bought both of them at the local Walmart. It's a little irritating to have one digital clock drifting like that and always being a little different than the rest in the house.
posted by WagonTyre at 6:07 AM on April 11, 2011