Power quality activities include the interaction of a source energy supply with load use demands.
In Part 1, we concentrated on what we called mystery events, or those disturbances that didn't correlate with any apparent electrical events on the user's electrical system or the utility's transmission and/or distribution system. In fact, Part 1's approach probably made you wonder if we were studying power quality at all, since we hardly mentioned the power issue. However, we did see a great many upsets to sensitive equipment and processes are caused by wiring and grounding conditions that lead to ground loops, "daisy chaining," and other ways of involving electrical noise patterns in our signal system.
The term "power quality," as used to denote our activities in this industry, is really a misnomer because the actual changes in the quality of electric power are very slight. Yes, there are events on the utility system that drastically affect sensitive equipment; for the most part, however, the powersupplies in use are built to protect themselves from all but the most excessive excursions in power.
As a result, a better name for our endeavors would be the "load sensitivity" or "load quality" business, and our task would be better described as the matching of a source energy supply with load use demands. This is a more accurate description of the challenge we continue to face in powering our loads.
Consider the operating voltage range of most power supplies: possibly from 6% or 7% higher than normal and from 12% to 4% lower than normal. In other words, the vendor of our sensitive electronic equipment is stating that its equipment will handle the stabilizing of the voltage once we deliver an operating voltage within the range listed. This range is normally well within that which a utility delivers. Even in the area of frequency, we find large utility systems are very stable at 60 Hz.
But the sensitivity levels of the new supplies do require us to prepare our source match-up carefully.
Let's look at those power disturbances that occur in line-to-line and line-to-neutral relationships. We call these disturbances normal mode or transverse mode because they originate from across or between the current carrying conductors of the system.
Here our problems are characterized by voltage sags or swells; high-speed line-to-line surges (steep wave front switching "spikes"); or outright power interruptions (sometimes momentary, sometimes for extended periods). We are looking at events that are closely related to observable conditions.
For example, we can tell when a large motor "sags" the voltage on a heavily loaded feeder and when a large "inrush" causes a similar type of disturbance to a piece of equipment nearby. We've seen the results of switching on the utility distribution system, particularly when capacitors are switched ON or OFF. In cases where shortages of energy are experienced on a distribution grid, our utility may have no other choice but to lower the supply system voltage so that it can provide the necessary energy to all users. This we know as a brownout and also by the term "rotating power reductions."
Probably one of the most frustrating of these events is the momentary interruption caused by a circuit breaker or recloser going through its operation, thereby protecting the system from a short-term electrical fault. Even when the system is restored in a comparatively short time frame (50 to 60 ms), the absence of power takes its toll on sensitive power supplies used in almost all modern control equipment.
Case studies in the plastics industry (those processes using microprocessor controls to operate a line of extruders, a forming operation, or a healing process) bear retelling at this point. Suddenly, the utility's circuit breaker goes through an operation, and the production line shuts down, most of the time with disastrous results.
Is there a major problem with the kilowatts of energy to the heaters, rollers, or forming devices? No. The problem is the sensitivity of a microchip in a controller section: This chip can't be without power for longer than 8 to 10 ms. At 50 to 60 ms, it sees a "loss of power" six to eight times longer in duration than it's capable of withstanding. As a result, it simply informs the machine to shut down. We might even say the chip sees a blackout; to its level of sensitivity, that's what it looks like.
So much for the symptoms; let's continue our site inspection survey.
Site inspection survey
As we begin our power examination, we want to know what events are occurring, and how the disturbances relate to a power event. Let's look at several examples for each type of event outlined above.
Voltage sagging or swelling. When we have events that correlate with the timing of special equipment functions or large load operation, we need to verify what range of voltage is actually available on the system. Could our troubled device be getting either too much or too little for some length of time, and thus shuttling down to "self-protect?" We remember the behavior of ferro-resonant devices when they must handle loads beyond their operating range: They go into a current limit mode, shutting down so as not to damage themselves.
When we see such an event, we should look for voltages that are "out of range." Many times, we can correct this problem with a change of taps on the transformer supplying the equipment. At least we can assess how to position the load in the middle of the range.
An extended sag is indication of a brownout and will have to be corrected with voltage regulating devices, where the condition is beyond the range of transformer tap changes.
Switching transients. Traveling "spikes" or impulses on the power line can cause problems for power supplies, especially those sensitive to sharp rates of change. A common disturbance on a distribution line is the switching of capacitors in or out of the circuit. This action creates a steep, ringing transient that can trick control systems into operation at the wrong time. One result is the tripping OFF of variable speed drives; the ringing raises the DC bus current level and increases the DC bus voltage above the "trip" threshold, thus causing a false shutdown. In cases such as this, a series reactor or choke would help calm the quick response of the traveling wave and "damp-out" its effect.
Interruption. Certainly we can identify with the complete loss of power for an extended period of time. Many of our sensitive processes must be able to count on a continuing source of energy. Here we see the need being handled by a UPS (uninterruptible power supply) and an engine-generator back-up. A form of stored energy, either in the battery bank or in the fuel tank or both, helps provide for this continuing power requirement.
Our site analysis should determine how critical the interruption problem is. Are the systems able to continue if the power is lost, perhaps because they have the stored energy within ? Or, will they shut OFF when power is removed? Can they withstand short outages, perhaps momentary circuit breaker operations? Or, is the sensitivity so high that even a cycle of energy loss will bring them down?
Many times a facility can profit by the use of two sources of power from the utility, and high-speed switching equipment to transfer between those sources in less than a quarter of an electrical cycle. Such a system avoids problems happening on one line by switching to the alternate feed so quickly that the load devices never see the disturbance on the preferred line.
When monitoring with a disturbance analyzer to capture the power events, remember that the placement of the monitor is critical to what you are measuring. If you're at the service entrance, you'll be seeing the utility system, but very little of the facility. As you move the monitor inward on the local system, you begin to see the addition of facility effects upon the power conditions. You may see what other loads do on the same bus, or how existing power conditioners behave.
We'll continue in Part 3 with more on site analysis, directing our attention to the growing problem of harmonic interaction.