Frequency power fluctuations

[markusjames]

Hello!
What's the implications on frequency fluctuations on 3 phase power?
From transformer to feed the lighting dimmer rack. We use Fluke power logger and recorded a frequency fluctuations from 50hz to 0hz in almost every 5 secs interval. It's random, different day and time. What may have caused this and possible solution? Thank you!

 

[Lucien Nunes]

A transformer cannot output power at 0Hz. If the supply is from the grid (i.e. not a stand-alone generator) then your frequency cannot vary independently. All consumers and generators on the grid operate at the same frequency which is controlled closely by the generation authority - if it varies by more than 1-2 Hz they have to shut down due to possible damage.

The plot showing variation between 49.92 and 50.04Hz represents normal operation of a power network. What appears to be the frequency dropping to zero is probably an artefact of another problem.

Can you give some info about the nature of the supply and distribution wiring up to the point at which the logger is attached?

 

[marconi]

To start off, I noted that when the Power Logger(PL) recorded 0Hz the reporting interval was every 5s, whereas when the power Logger recorded 50Hz the reporting interval was every 10s.

I also looked at page 14 of:

http://assets.fluke.com/manuals/PowerLogumeng0100.pdf

The Fluke PL calculates the True Definition of Voltage Unbalance, viz the ratio of the negative sequence voltage to the positive sequence voltage = Vneg/Vpos. This is called the Voltage Unbalance Factor as defined in IEC 61000-4-30. There is also a simpler less accurate NEMA method which does not take into account phase angles - see for a description of both:

Voltage Unbalance – Voltage Disturbance

From the voltage and current inputs, the FPL analysis the unbalanced voltages into the two balanced positive and negative sequences plus the zero sequence. If the system is balanced then there is no negative nor zero sequence. A good measure of the percentage of unbalance is then the ratio Vneg/Vpos.

I think - don't know for certain - that when it is selected that the Fluke PL uses the negative sequence waveform eg V1N, it uses it to count zero crossings over a timed interval to then calculate frequency. So, for periods when the system is balanced the negative sequence waveform will be of low amplitude and noise will confuse/interfere with the electronics ability to detect zero-crossings sometimes stopping it taking place - zero frequency is output. Again, I don't know for certain - but I reckon that over the longer interval (10s) noise may be less of a consistent problem and, it may be less likely that the zero sequence is of such a small amplitude throughout the interval for noise to be troublesome and stop frequency estimation.

My wife has just returned home from hospital so I must cut short by saying that it could then be a question of understanding how the Fluke PL works, calculates and displays information to explain the quirky drop to 0Hz every so often. If for example V1P for Positive Seq ( cf V1N for Negative Seq) was selected instead, recorded and displayed I suspect the frequency would not dip to 0Hz because it is always of a decent amplitude for noise not to be a problem.

Or something like this..others can expand or correct.

 

[marconi]

I suppose now my theory in #17 is wrong.

Studying your image 2 in #12 it does not look exactly like the one illustrated on page 14 of

http://assets.fluke.com/manuals/PowerLogumeng0100.pdf

for when the Frequency/Unbalance tab is selected. So your data logger does not appear to me to have the True Unbalance capability. Instead it records and displays max, min and average Voltages and Currents. The reference did say some Fluke DL's don't analyse Unbalance.