Before moving equipment into a new or renovated facility, establishing baselines of the quality of the supply and comparing it with the susceptibilities of the equipment that will be going into the facility is a good first step.
Monitoring for one business cycle, as well as obtaining reliability data from the electric utility, is key to seeing how often the equipment will typically be subjected to various power quality disturbances. A business cycle is based on how the facility operates. In many applications, the operation of the facility is different on Saturday morning than on Wednesday afternoon, so a one-week monitoring period is recommended.
Of course, the reliability of the electric network usually varies during different times of the year (e.g., lightning versus ice storms), but no one is going to wait that long before starting up the facility.
IEEE Std 1366 Guide for Electric Power Distribution Reliability Indices is a good place to find more details of the monthly indices that most utilities are required to track by their Public Utility Commission, such as SAIFI, SAIDI and CAITI. While there are many indices in use today, the primary indices being used by most utilities are SAIDI (average amount of time a customer would expect to be without service), SAIFI (average number of times a customer would expect to see an interruption of more than five minutes) and CAIDI (average duration of an interruption).
While there has been some work in the standards-making groups (NEMA, IEEE, IEC, etc.) to require susceptibility and emissions data in the spec sheets of equipment, such as adjustable speed drives, the pickings are still slim.
It may require contacting the vendor directly or running some tests of your own. You might find the results quite different from what you would expect. The programmable logic controller (PLC) may ride through sags that are much deeper and longer in duration that those that will cause the contractor to drop out.
Another IEEE document, IEEE Std 1346 Electric Power System Compatibility with Electronic Process Equipment, has a good way to compare the quality of the supply with the susceptibilities, overlaying them on the same graph to determine vulnerability. Other quality of supply graphs, such as the CBEMA and ITI Curve, are well known, but are general purpose and may not apply to the equipment in your facility.
The steps with an existing facility are similar to the new construction; however, you often have less freedom to run tests with the equipment, as few facilities can afford downtime in today's global economy. In addition, having PQ phenomena-resulting from the equipment and electrical distribution system within the facility-clouding the picture complicates monitoring data.
Just because you recorded a sag at the service entrance doesn't mean that it originated from the electric utility. A large motor start or even the turn-on of a copier can generate such. There are a few commercially available PQ monitors and software that can determine the direction of the sag in most circumstances, based on the same rules of Kirchoff and Ohm that have been discussed in previous columns. In general, if the current goes up significantly when the voltage goes down, then the problem probably originates downstream from the monitoring point; or else, the origin is probably upstream or back toward the utility when the current does not go up significantly when the voltage goes down (see Figure 2).
There have been some publications that claim that the quality of the supply is deteriorating as result of deregulation and distributed generation, though there has not been a nationwide benchmark program in the United States since the early 1990s with which to validate such claims. Diversion of funds at the electric utilities from O&M tasks, such as trimming the vegetation away from distribution and transmission lines, does occur to improve the bottom line.
In fact, such an encroachment is cited in the reports on the blackout of Aug. 14, 2003. However, a significant portion of the distributed generation equipment is used to take loads off the power grid in times of diminished grid capacity, so there hasn't been much public documentation on the interaction of this affecting the reliability of the grid in practice.
Windmills are said to induce flicker in parts of northern Germany, but overall, I believe that the jury is still out. And with the series of interconnection standards under IEEE Std 1547 Standard for Interconnecting Distributed Resources with Electric Power Systems, hopefully, only the positive aspects will be noted in reliability going forward.
Once you know what needs fixing, you can economically go about it. Just putting in more mitigation equipment isn't always the best solution. Mitigation equipment, such as any piece of equipment, has a mean-time-between-failure (MTBF).
A story has been circulating about a semiconductor facility that put in hundreds and hundreds of UPS units on all of the lines, even where there weren't significant problems reported. The result was that the lines were down more often due to the normal MTBF of the thousand units, rather than what was experienced with the reliability level of the electric supply before.
Another classic is the proliferation of TVSS or surge suppression strips to improve the power quality. While a properly sized and installed TVSS can reduce voltage transients, they do not eliminate harmonic distortion nor provide the energy to fill in the voltage waveform on the most common type of PQ event-the sag.
Each type of PQ phenomena has its recommended solution, such as filters for harmonics and flicker, UPS units for sags and swells, and even quarter cycle solid-state transfer switches to transfer the incoming supply from one distribution feeder to another before the “blink” can affect any equipment.
You also need to be sure that the mitigation equipment is compatible with other mitigation equipment. In high reliability or continuous uptime facilities, the multiple levels of mitigation and backup equipment can make for challenging troubleshooting when something does go wrong.
Minimizing the effects of power quality phenomena on the quality of the process, whether in data centers, the factory floor or even a guided missile cruiser, is not a one-time adventure. Though most facilities still use handheld or portable monitors for troubleshooting when problems occur, this type of after-the-fact, forensic process is often time consuming and full of guess-work, because the troublesome event may not occur again for weeks.
Handheld/portable PQ monitors range in cost from a few thousand to more than 10,000 for the more elaborate, full-spectrum instruments. Permanently installed monitors also come in a wide range, from revenue meters with limited PQ to PQ-centric LAN-based systems that provide data, information and answers through the Internet in seconds. This usually aids in getting the facility back in full production as soon as possible, maximizing the bottom line.
Power quality is sort of like your interaction with the medical profession over the course of a life. From birth to the growing stages to the retirement days, a constant checkup and the right medicine at the right time will go a long way to a happy, healthy life. EC
BINGHAM, a contributing editor for power quality, can be reached at 732.287.3680