When using a power quality monitoring instrument, there are some “gotchas” to watch out for; otherwise, your efforts can either be wasted or, perhaps worse, misleading by coming to conclusions that aren’t valid. The complete list is much longer than the one below, but these are among the top of the hit parade.
Measuring with equipment not suitable for what one wants to observe
Though it can be a daunting task, a careful read and reread of the power quality monitor specifications is well worth it. Go beyond what is listed on the product website or brochure. The user guide or operating manual is more revealing if read with a “I’m from Missouri” mind-set. (Lest I offend anyone, the intent is to piggyback on the state nickname as the Show-Me State.) Look out for things that are more marketing triumphs than scientific facts. Phrases such as “designed to work according to a particular standard” is far different from “fully complies with and certified.” If you want to capture high-frequency impulsive transients, a sampling rate of 128 samples per cycle isn’t going to cut it. Likewise, if you have a 24-pulse converter in an adjustable speed drive front-end, you aren’t going to be able to measure more than a few harmonics that it generates. An alternating-current-coupled voltage input and virtually any non-Hall-effect current probe is useless in determining direct-current offsets.
Measuring with equipment that isn’t measuring what is really occurring to prove what is desired rather than reality
A caller on a radio talk show commented how a particular group of political pollsters will either throw out polls that don’t give them the answers they want or modify the questions in a way that the reader has to support their position. This method is sometimes used by a person setting up the monitor to support the outcome that they want or, as above, using a monitor that can’t measure a particular phenomenon. Saying that the transient system works fine because of the lack of recorded transients during a thunderstorm isn’t surprising, for most power quality monitors on the market never do. Having software that does a one-minute temporal aggregation on sags can significantly reduce the count. Even though the sensors on the process-control system malfunctioned many more times doesn’t mean that they aren’t the source of the problem.
A slight twist on this comes from a person who wanted to measure the harmonics on the output of a pulse-width-modulated adjustable speed drive but knew that such a waveform would be impossible for the phase-locked-loop circuitry of the digitizing circuitry to synchronize to. So, he made a filter that gave a nice clean sine wave to the voltage inputs. To his disappointment, the instrument didn’t show any significant harmonic content as he varied the frequency of the drive. His filter worked just fine in doing that.
Measuring with equipment that is influencing what is being measured
Though the Heisenberg Uncertainty Principle is usually referred to in the quantum mechanics realm, it also provides a caution for power quality monitoring. It has been paraphrased as, “the observer becomes part of the observed system.” Most voltage input and current-measuring devices in the newer power quality monitors have high enough impedance so as to not load down the circuitry to which they are connected. But, these instruments often are connected to communication systems, such as local-area networks, through cables that can either cause ground loops, serve as antennas for receiving or transmitting noise, or provide a vulnerable path to the “innards” of the monitor if lightning should couple into the communications system. Most monitor power supply inputs are protected with transient-voltage surge suppressors that would clamp any transients if the monitor’s power supply were plugged into the same circuit that it was monitoring, such as those with “phase-powered power supply” specifications.
Measuring in the wrong spot for what you are looking to determine
Though this one seems like a no-brainer, it still comes up when people don’t take time to determine where all those wires really go in a panel or rely on an outdated, one-line diagram. A company had built a case for “improper power” being supplied by the electric utility from a monitor that wasn’t connected at the point of common coupling, where the utility meets the customer. It was downstream off a branch circuit toward some of the customer loads. This means that the algorithms commonly used to determine the direction of a power quality disturbance (source/upstream or load/downstream) couldn’t be applied. Hence, there was no way to prove if the cause of the problem was another part of the facility or the utility. Another common problem is not matching the current probes up with the proper voltage channels. The watt calculation is done on most instruments with every sample of each taken. Those data points aren’t saved because there are too many, so there is no way to go back afterward and create the proper power data.
Though these gotchas may seem like common sense, they happen far too often to be overlooked. Sometimes common sense isn’t too common.