Recently, a PQ consulting expert sent me some interesting data from a facility he monitored. The data was the result of periodic PQ monitoring that is conducted at the facility’s service entrance. There wasn’t a report of problems in the facility, though the resulting data shows several potential problems that may cross that “threshold of pain” in the future. It also illustrates the value of being proactive, rather than trying to solve the problem in the forensic mode, after the fact.
RMS variations—also known as sags, swells and interruptions—apply to currents, though most people associate them only with voltage. Having a PQ analyzer that can trigger on current as well as voltage is very valuable in finding out the “whodunnit.”
The first clue in the data (see figure below) was the repetitive nature of the swells, in magnitude, duration and time interval between swells, being approximately 43A, 2 cycles, and 50 seconds, respectively. It indicates a single load causing the swells that has some sort of timer mechanism in its operation.
The second clue is in the waveform that is shown in the table above. The Channel C current waveform is shown increasing about 2:1 with a very small phase shift between the current and the voltage (not shown here).
It does have some distortion visible near the peak at the 90 and 270 degree points on the wave, so it is not a pure sinusoid with fundamental frequency current only. There is also an initial DC offset and 2 Hz decaying oscillation in the envelope of the waveform observed. That usually means that is a combination of a resistive and electronic load with a rectified input, switching power supply as being source of the current swell. Rectified inputs draw current based on the load levels in a nonlinear manner; that is, they don’t conduct current during the entire cycle, so it isn’t a sinusoidal current waveform anymore-—the classic source of harmonics.
This is a fairly common signature of a certain type of loads that have both rectified input power supplies and resistive heating elements in them, such as laser printers and copiers. Turns out it was a copy machine, which has a heating element that cycles on periodically, about every 50 seconds if not in the power-save mode.
The good news is that the source impedance of the service entrance is very low, also known as a “stiff source.” The 30A change in current resulted in approximately a 0.5V drop in the voltage on Phase C. The next test would be to determine the source impedance as seen from load, or, looking back electrically towards the service entrance. This is determined by computing the change in voltage divided by the change in current at a number of different points in time.
Another data observation is the significant current imbalance on each of the phase conductors, which contributes to the large current recorded on Channel D, which was monitoring the “neutral” or grounded conductor of the 4-wire wye circuit at the service entrance. The Channel D current levels ranged from 10–30A RMS, with 3A or so consisting of harmonic currents.
Here again, the stiff source has been beneficial, as the neutral-to-GND voltage is less than 0.2V, and the voltage imbalance was averaging 0.3 percent and a maximum of 0.5 percent. In general, the currents in each of the phase conductors should be much closer in value to each other.
In fact, the neutral conductor is carrying more current than Phase B conductor. Loads could be shifted from one of the heavily loaded phases to a lightly loaded phase. Whereas these levels currently aren’t at the fire-hazard level, neutral conductors have been known to “smoke,” especially in older facilities where they were originally allowed to be undersized compared to the phase conductors.
The data yields one final clue that could have impacted much of the aforementioned observations—the time of day. Facilities have different load profiles during different times of the day, days of the week, and even months of the year. Monitoring should usually be conducted over one business cycle (often a week), which takes into account those factors.
All in all, this PQ checkup shows that things at this site are generally pretty “swell,” withstanding, for now, the effects of the copy machine on the electrical system. This, unfortunately, is not the norm today. Proactive PQ monitoring should be an integral part of all facility maintenance programs. EC
BINGHAM, a contributing editor for power quality, can be reached at 732.287.3680.