At a recent training session in Washington, DC, we plugged a power quality monitor into a baseboard outlet to get data to illustrate the different parameters and characteristics of voltage and current waveforms. As is true in most office complexes, the top of the voltage waveform was flat topped—a sort-of “Mount St. Helens effect.”
Correspondingly, the voltage harmonics consisted of 3rd, 5th, 7th and 8th, as one would expect when there are a significant number of single-phase loads with rectified inputs that primarily draw current in the middle of the cycle, when the voltage is at its peak value (90 and 270 degrees). The voltage drop in the source impedances (wiring, transformers, protective devices) occurs only in the middle of the cycles, leaving less voltage for the loads causing the flat-topping.
At this site, the amount of voltage reduction on the peaks of the voltage waveform was not severe. However, it still means there is less voltage there when a sag occurs, giving a little less margin from the point of susceptibility for the connected loads. But the tenants had no idea such an event was happening.
Then the monitor beeped a couple of times, the indication that power quality events or disturbances had occurred. Though it disrupted the planned schedule of the session, curiosity got the best of us, and we hit the button to display the events. There were several transients, with the most interesting being a unipolar positive transient—a sudden increase in the voltage, very short in duration, in just one direction away from the normal sine wave curve, one that adds energy to the waveform (see Figure 1).
Zooming into the waveform shows the peak voltage was just over 300V, with a pulse width of around 13 usec. (Figure 2). The clipping of the top of the transient stood out. Given its fast rise time (faster than 30V/usec) and the abrupt clamp at around 300V, the surge suppressor strip where the computers (same ones causing the flat-topping) were plugged into most likely did their job and prevented a much larger transient from propagating throughout the facility.
Whereas there is no way to exactly determine what that peak voltage would have been without the surge protectors, the rise time and pulse width indicate it likely would have been in excess of 500V. If this wasn’t enough to damage the equipment plugged in and operational at the time, it likely would weaken the semiconductors and capacitors that experienced the effect of such a high energy transient. And even equipment that is turned off but has some portion running could be damaged, such as a television set that turns on with a remote control.
The monitoring period didn’t continue much longer before several more transients occurred. This also takes its toll on the surge suppressors, which give a little of themselves each time they have to absorb surges; this can cause them to clamp at a higher voltage or fail completely in the future if such abuse continues. But the tenants of the building had no prior idea that this was going on within their walls.
This changed the awareness of the tenants as well as the focus of the training. “What caused it, and where did it come from?” is the typical question from those suddenly enlightened about events they didn’t see or were unaware of before. To determine whether it came from within the facility or from the electric utility distribution system, one normally needs to monitor the voltage and current, starting at the point of common coupling, where the utility power meets the facility. Unfortunately, no one had the proper personal protective equipment to be able to safely connect at the breaker panel. Some countries have access ports to do this safely without removing the cover but not here in the United States. Since there were no thunderstorms in the area that day, we ruled out lightning as the source, but it would have to be some sort of source with the ability to add significant energy to the voltage waveform.
This incident is the opposite of another transient recorded, which was a negative transient—one in which energy has been removed from the voltage waveform. That type of transient is more likely to have occurred by a load within the facility that turned on and drew a very large amount of current for a very short duration, such as charging a large capacitor. Again, it would take a tour throughout the facility to note the different equipment in the building in addition to more monitoring at different locations within the facility to narrow down the source. This is not the instant fix the tenants were hoping for.
Such is a typical day in the business of power quality monitoring for those who have the tools to know what lurks within in the walls. EC
BINGHAM, a contributing editor for power quality, can be reached at 732.287.3680.