Frequently, power quality (PQ) phenomena that result in changes to the voltage waveforms aren’t noticeable without specialized instruments. For example, a power factor capacitor switching transient isn’t likely to make the lights dim or flicker, yet it may cause the adjustable speed drive (ASD) to trip off line. The notching of the voltage waveform during the commutation period of the solid state rectifiers that make up the AC/DC power conversion section of most three-phase “electronic” loads, such as ASDs, may eventually damage electronic components or the windings of the motors that they are connected to. When the resulting damage becomes catastrophic and something fails, it will obviously be noticeable, but not the events that lead up to the failure. However, there are many times when observations by humans become as valuable as having a power quality monitor on-site.
At a manufacturing facility, there was a rapid series of the lights dimming, followed by the e-mail system going down. Whereas the second problem was the result of a failure to replace the aged batteries in the uninterrruptible power supply (UPS), the cause of the series of voltage sags were determined by auditory data—hearing a jackhammer being used in the immediate vicinity of the building.
The good news is that the jackhammer did not hit a hidden power cable. What it did show was that there was a loose electrical connection in the facility that required tightening. In similar incidents at other facilities, vibration from passing trucks or a pump motor starting up on the floor below resulted in similar events.
Those two incidents required much keener hearing to make the correlation than a jackhammer outside someone’s window. But, at least someone at the facility was able to make the connection that the dimming lights occurred at the same instance as the jackhammering.
The incident is a reminder that safety comes first. Always check for buried utilities; never assume that there aren’t any. A number of years ago, Newark Liberty International Airport, one of the world’s busiest airports in the New York City area, was shut down for several hours when a pile driver missed the mark and drove a pile through one of distribution voltage feeders to the airport. Fortunately, no one was hurt, but there were a lot of angry travelers.
Sometimes the correlating event isn’t a mechanical one, but an electromagnetic-based event. In a computer room of a data center, there was a problem with the system crashing at almost exactly 8 p.m. in the evening. At that time, the only employee in the area was the data center system operator, who claimed to just be sitting there, not touching anything when it happened.
After a number of incidents and nothing observed on the power monitoring system, a power quality monitor with a radio frequency (RF) detector module was brought to look for something coming through the air rather than the wires. Sure enough, at around 8 p.m., there was a large burst of radio frequency (RF) recorded.
It turned out that the security guard was making his rounds through the facility and would pass outside the computer room at 8 p.m. each evening, at which time he would key his radio to report to his supervisor. Improperly grounded communication cables between peripherals and the mainframe computer became the path for the RF to get into the computer system and wreak havoc.
In a commercial facility, the accounting manager’s computer would restart in the middle of doing the morning cash flow analysis. Thinking that something was wrong with the hardware or the software, the IT department was contacted to run tests. After they could find nothing and the problem persisted, eventually the maintenance department was contacted. Their first tool was a typical digital volt meter (DVM), which updates its display about 4 times a second. The meter showed a nearly steady 122 volts, so the verdict was that it wasn’t the power.
DVMs are good tools for some things, but they aren’t very good tools for detecting intermittent disturbances, unless you get lucky. Eventually, a power quality monitor was connected at the outlet where the accountant’s PC was plugged in, and a severe sag was recorded at the time the computer restarted.
Once again, the correlation came down to someone else in the area doing something—the accounting clerk turning on the large copy machine. The inrush current from energization of the copy machine caused a sag low enough to cause a misoperation of the power supply in the PC, with the resulting restart. Relocating the PC to a different circuit was the quick solution.
Finding the source of the problem with just the power quality data and without the human observations would have been a much more difficult, if not nearly impossible, task. Encourage people in the facility to be observant and write down the time when things go astray, what they were doing, and any other visual or auditory observations that might be able to correlate to the source of the PQ events. EC
BINGHAM, a contributing editor for power quality, can be reached at 732.287.3680.