I was recently called by a homeowner to investigate water dripping through the ceiling. Water through a ceiling is surprisingly not uncommon, from slow drips to entire ceilings crashing down. As with most investigations, the first question from the homeowner is “Where is it coming from?” Also, the facts as relayed by the homeowner need to be carefully vetted. In every case, one thing holds true—water flows down from the source, unless it is a high-pressure leak pointing up. The hunt starts from where the leak is visible and works its way up to the highest point where unwanted water is found. In this case, it was a leaky diverter in the shower above the stairwell ceiling.
Power quality troubleshooting investigations to find what caused the process to be interrupted also has its own indisputable rules. Ohm’s and Kirchhoff’s laws define the relationships between voltage and current. Regardless of whether the problem was caused by harmonics, voltage sags, transients or any other PQ phenomena, what happens to the voltage is dependent on the current and the appropriate impedance.
The term “appropriate” is also a key factor here because source, wiring and load impedances for a 1 MHz transient are not the same as for the 13th harmonic of the power frequency 60 Hz. It is referred to as Z for impedance, not just R for resistance, because most electrical systems have capacitance and inductance values. Resistance is a constant impedance over the frequency range, but the impedance of a capacitor goes down as the frequency goes up, while it goes up for increasing frequency for an inductor. Voltage (V) equals I (current) multiplied by Z (impedance). The sum of the voltages around a closed circuit equals zero, as does the sum of all currents at a node.
With that quick review of the rules, let’s follow the steps for a typical PQ troubleshooting scenario. Most requests for an investigation don’t happen after just one occurrence of the process interruption— unless it resulted in a large monetary loss Having more data points is helpful for the investigation, provided the customer took accurate note of what occurred, when it occurred and what was affected. Most requests also don’t occur at facilities where there is already a power quality monitor(s) in place.
Example 1: Overloaded receptacles
The sales area on the second floor of a manufacturing company had the PCs in one cubicle locking up or rebooting several times a week. No other area in the building reported similar problems. No lights were known to blink when it happened, nor were the PC users doing anything different. After blaming faulty PCs and software, a PQ monitor was connected to the same receptacle as one of the problem computers and at the distribution panel where the circuit originated. The latter saw nothing unusual; the former recorded a voltage sag below 80% of nominal at the same time as the PC problem.
A closer look at what was plugged into the receptacles at the base of the cubicle found several computers, a heater (though turned off), a coffee pot and a laser printer. Examination of the current waveforms showed the typical heating element pattern for the coffee pot and the laser printer warming up before printing. When both of those current increases occurred at the same time, the voltage drop in the smaller-than-14 AWG wiring of the cubicles and the connectors between each panel resulted in the voltage drop below the threshold that the computer’s power supply required to remain stable.
Example 2: Capacitor failure
In another example, a large bulk mail facility had capacitors on their own electrical substation blowing up once every couple of months. The company had recently gone through a lighting and motor upgrade in the facility to significantly reduce the electrical consumption. Workers were baffled as to why a smaller load should cause the capacitors to destruct. The usual monitoring locations where the problem occurred and back toward the source weren’t needed because the capacitors were at the power source. The voltage waveforms revealed significant 7th harmonic current waveforms superimposed over the 60 Hz current waveforms whenever the capacitor banks were switched on by the daily timer, which was used to improve the power factor during peak operating load.
Large currents result in significant voltages because the harmonic impedance at this frequency was actually a resonant point. No one had recalculated the amount of capacitance needed after the overall load current was reduced by their improvements. The circuit now had a resonant point at the 7th harmonic—a typical current for the newly installed adjustable speed drivers. Continual overvoltage resulted in capacitor failure.
Next month, I will review a few more typical troubleshooting scenarios, along with a couple where the rules are applied in an atypical manner.