From David Letterman to the FBI to Billboard magazine, the “Top 10” can be applied to just about any category, including power quality phenomena. Although it is a highly subjective process, there is some merit in being more than just familiar with them. If you are able to quickly recognize the signatures of the Top 10 RMS variation, transient, harmonic and other PQ events, you will have covered the majority of disturbances that you are likely to encounter while monitoring at an industrial, commercial or residential facility. We will deal with RMS variations first, as they are usually the most common disturbances on electrical systems.
RMS variations are changes in the voltage from the steady-state nominal value that last a cycle or longer. Typically, if the voltage drops to between 90–10 percent of nominal, it is called a sag, while 110–180 percent of nominal is a swell, and loss of voltage completely (below 10 percent) is an interruption. While the voltage RMS and waveform graphs alone can be used to identify the source of the disturbance, adding the current graphs makes it much easier.
Utility or source-side RMS variations are usually the result of a fault occurring on the distribution system and causing the utility’s protection system to take action to protect itself. This is evident in the first phenomenon on the list, the “bathtub sag” in Figure 1.
The name comes from the “squareness” of the waveform and the RMS plots. The voltage reduces in a step change, stays fairly constant, then returns back up as a step. It is usually 5–10 cycles long and 70–90 percent of nominal. A fault—such as a downed wire—causes it, drawing a large amount of current. The breaker in the substation will wait a number of cycles of overcurrent to see if the fault self-clears (such as a tree branch hitting a conductor) before opening to try to clear the fault. If you are located on a parallel feeder, you see the sag. Downstream from the breaker, you will see an interruption following the sag when the breaker operates, such as in Figure 2.
If the fault or overcurrent condition still exists when the breaker recloses, it will try several times to clear the fault before it “locks out,” requiring human intervention to restore power. This recloser sequence of RMS variations depends on the utility, but it often is three to 10 attempts about a second or so apart.
A severe system-wide disturbance can result in sags and interruptions for those on the downstream side of the affected substation(s). If enough system capacity is affected, such as during the August 2003 Northeastern U.S. blackout, some locations will experience a swell or increase in voltage, as there is less load on the remaining generating capacity, as shown in Figure 3.
A swell can also occur on a smaller scale when a large load on a particular feeder or substation suddenly goes offline. The voltage in a local area is regulated by tap-changing transformers that can adjust the voltage by switching to a different tap that has a different turns ratio. So as to not react to very short-duration load changes, the tap changer may wait a minute or so before doing the switch. Since less current is being drawn on the circuit from the load going offline, there is less voltage drop in the line, and the voltage will swell temporarily until corrected.
Load-induced sags within a facility are usually more common than utility side. Several of them also have distinct signatures, based on how the load uses the current. Among the most common of these are motor starts. When an AC motor starts, there is an inrush of current to magnetize the windings and get the motor turning. The larger the horsepower, the larger the inrush current. More current means more voltage losses, so less voltage is available for other loads on the circuit. Figure 4 is a small horsepower (hp) motor, and you can clearly see the inrush current abruptly starting and then decreasing in an exponential-type curve.
In the larger horsepower motor start, the duration is longer and the voltage sag is more severe. The RMS plot clearly shows the exponential shape of the voltage recovering.
Another common load that can cause short duration sags are heating elements, found in coffee pots, laser printers, HVAC loads, refrigerated vending machines, and other equipment that cycles repetitively. To determine the source, it is important to look at depth and duration of the sag, the repetitive pattern, when it occurs (e.g., 8 to 10 a.m. in an office environment might be the coffee-maker), as well as the power factor and current harmonics.
The sags in Figure 5 were found to be from a laser printer, where the heating element came on periodically every 45 seconds while it was idle and the current waveform suggested a resistive load. But the current waveform (not shown here) before and after looked like a typical single-phase, rectified load of a switching power supply found in most computer and IT equipment. Note that in single-phase circuits, the neutral-to-ground voltage mimics the current.
Some other sags in the Top 10, but not illustrated here, are a sag on one phase and swell on other two from high impedance grounds; long-term, sustained voltage reductions; and sags from the capacitor energization when plugging in nonlinear load. Different types of manufacturing operations will also yield identifiable signatures from their processes (such as the melting stage of an arc furnace).
Being familiar with both how the local utility system operates as well as operations within the facility can be quite helpful in creating your own Top 10 list of RMS variations.