PMU for Event Analysis

[111R]

Most microprocessor relays are capable of sampling at rates of 128 samples per cycle or more. Would it ever be possible to get this type of resolution with Phasor Measurement Units or will these always be more for longer duration event analysis such as power swings and such?

The highest sample rate is currently 60 samples per second.

Thanks

 

[davidbeach]

I've heard rumors of 120 sample per second PMU data, but as a continuous stream of data there's no reason to have more. High speed recording should be limited to interesting bits of time or all you'll ever see is the hay-stack and no clue about the needles. High speed recording of very slowing changing things doesn't seem to be worth the data storage requirements.

 

[xnuke]

There is a difference between PMU sample rate, which can be up to 48 samples per cycle (2880 samples per second), and PMU data rate for archiving and transmission, which I've also heard may be up to 120 samples per second these days. Obviously, downsampling occurs.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[111R]

OK, I think that's where I'm confused. I wasn't sure if the 60 samples/second was a sample rate or a rate at which data blocks (with multiple samples within) were sent.

 

[davidbeach]

The PMU data coming out of the relay is derived from the same data stream as the protection data, many times per cycle. A "slow" relay would be running at 960 samples per second; that's way too much data for anything you'd do with PMU data. For the fault event files, that data is continuously going into a buffer and if an event is triggered a block of data is stored in memory. The newer relays relays may be sampling at 8kHz or higher and allow the option to save event files with fewer points, perhaps 1, 2, 4, or 8kHz so that more and/or longer events can be held in the relay's memory before they begin to be overwritten. I'm aware of at least one relay that samples at 1Mhz, data from that one that hasn't been significantly down-sampled is drinking from a fire hose.