There is a saying that, when you point your finger, three fingers point back. When a facility encounters a misoperation that the manager thinks was caused by the electric power, the first reaction is to point the finger at the utility.
With the oft-quoted statistics saying that the majority of power quality problems originate within a facility, the utility is quick to point the finger back.

I recently encountered a case with a lot of finger pointing. Yet, no one was looking at the years of power quality data that had been collected. Granted, the monitoring system installed was a single-point in a multibuilding campus that was fed from two separate distribution sources, so it showed just a fraction of the total picture. But, they had answers that no one used when making their accusations.

The rules of thumb for determining the origin or directivity of a power quality disturbance are pretty straightforward and based on the two laws for solving power quality problems—Ohm’s and Kirchhoff’s Laws. In quick review, voltage equals current times impedance, and the sum of the voltages around a closed circuit or the currents into a node equal zero. Directivity is with regard to the source of the disturbance, which can either originate upstream toward the source or downstream toward the loads.

Probably the easiest rule of thumb for the directivity of a sag is, if the current increases significantly when the voltage decreases, the origin is most likely downstream from the monitoring point. Like the imprecision of measuring with a thumb, the meaning of the word “significantly” can vary by quite a bit, depending on the source and load impedances. But generally, the increase in current will be three times or more from the steady-state condition prior to the sag beginning. This is particularly true if the source of the disturbance is the startup of a large horsepower motor.

The corollary to the downstream rule is a bit more involved, as the predominant type of load on the monitored circuit can change the resulting voltage and current waveforms. If the loads are mostly linear, the current will decrease approximately in proportion to the decrease in voltage during the sag. A sag to 80 percent of nominal would result in the current during the sag reducing 20 percent also. If the loads are constant power devices, the current will increase proportionally to make up the difference in the reduction of voltage so that watts remain constant, given the equation of watts equal to voltage times current times power factor. The third and probably most common load nowadays is a nonlinear or electronic or rectified-input-switch-mode-power supply type load found in most information technology equipment, adjustable speed drives, consumer electronics, etc. Since the first stage of the power supply are rectifiers that charge up the direct current (DC) voltage of a capacitor, the drop in alternating current (AC) voltage below the DC voltage will cause no current to flow into the power supply. This will eventually change as the energy is taken out of the capacitor to feed the rest of the power supply circuit and keep the load running, but initially, the current will reduce significantly and may even go to zero. So, for an upstream-originated sag rule, the current will reduce somewhat, increase somewhat or reduce to near zero initially during the sag.

Of course, three-phase loads with a sag on one phase only can complicate the picture a bit, but one can still usually determine the directivity of its origin and, sometimes, even the cause of the sag. The figure above shows such a case. Before the sag officially occurred near the peak of the end of the second cycle of voltage L3-N (blue waveform), the voltage of L3-N and L2-N were already slightly reduced and L1-N slightly increased. This is a precursor to signify something is wrong, which finally occurs with L3-N reducing to near zero volts very abruptly, then recovering to about 25 percent of nominal for 4.5 cycles before it returns to within 10 percent of nominal.

At the same time, the current decreases within a cycle to approximately 25 percent of the presag value of L3. L1 and L2 also reduce, though not as much. This most likely indicates that the sag caused some loads to trip offline. When the sag ends, the current levels of all three increase to approximately the same (though L3 has a larger value for the first cycle). But then three cycles later, they all increase again, probably due to a restarting load.

BINGHAM, a contributing editor for power quality, can be reached at 732.287.3680.