IEEE1346-1998: A New Tool To Combat Voltage Sags

Mar 1, 1999 12:00 PM, By R. Larry Morgan, P.E.

Recently approved, this new standard provides a method of determining compatibility between utility-caused voltage sags and your electronic devices. It also provides guidelines for corrective action.

Downtime and lost money: That's what many companies face when their electronic process control devices experience voltage sags. Recognizing this common problem, IEEE developed a tool you can use to minimize such occurrences: IEEE1346-1998, Recommended Practice for Evaluating Electric Power System Compatibility with Electronic Process Equipment. You can find information on its availability through the www.standards.ieee.org website.

The standard also addresses the age-old quandary: "I know how to solve this problem, but the money isn't available. This is a low bid job, I'm on a budget, and I can't deviate from established goals."

In response, the standard provides information on how you can determine a statistically valid up-front cost impact of plant downtime for your site and equipment.

Predicting a cost impact. How do you go about this? The answer has two parts:

  • By predicting a statistically valid rate of occurrence, magnitude, and duration of each voltage sag for your plant site over a period of one year; and
  • By characterizing the susceptibility of your sensitive electronic devices to such electrical events.

Basically, you put this data into a financial format. Then compare the up-front cost of installing equipment that will allow your electronic devices to ride through voltage sags with the cost of inevitable plant downtime. Management can use this tool to determine the best course of action.

A look at the details. First, turn the utility event into a statistical plot. (See Fig. 1, original article) Note: The plot is a hypothetical site, served power by a local utility. All events in the light blue area have an equal chance of occurrence. A voltage sag down to 80% of normal and lasting 20 cycles has the same chance of occurring as one down to 30% and lasting two cycles. Also, consider the majority of events are over in 15 cycles.

Likewise, you can characterize sensitive electronic equipment by seeing how low a voltage it will operate on, as well as how long it can ride through an electrical disturbance. You can identify these characteristics in the susceptibility chart. (See Fig. 2, original article.) Any event longer and deeper than the dropout box boundaries will inevitably cause problems. Now, overlay the two curves, as shown in the Fig. 3, on page 33, original article.

The star indicates voltage sags will cause the sample 24V (output) power supply serving a critical control device to produce five to 10 process shutdowns per site per year.

The next step is to predict the cost of the process interruptions for the year and end up with the cost of using this power supply in your process equipment. You can then do an analysis of the options available to alleviate this problem and evaluate the associated investment choices.

Once you can predict the cost of an interruption to your process, you can determine what investment options are best for voltage sag ride-through. To find the probability of a sag's impact on your equipment, you should request voltage-sag characteristic plots for your site from your utility. Also, ask the manufacturer for plots on its equipment's susceptibility to sags. After adjusting the two charts so they have the same scale, overlay one over the other to observe where impacts occur.

Getting voltage sag activity. Your utility may not have voltage sag information, especially for specific areas of its service territory. Instead, you may have to rely on sag information from actual measurements, predicted performance, or "typical" data. Nevertheless, you'll need reasonably accurate data for a good financial analysis.

Actual measured voltage information for the facility gathered over several years is accurate, so long as the supply system won't change significantly in the future. However, some utility engineers believe several years can be too short. Even specific data gathered by the utility varies somewhat for each year, each season, and each location. Voltage level and other factors will also influence data. Making projections from history may not be any more accurate than typical data, since many utilities frequently reconfigure distribution circuits.

Where to measure sags is important. Utility data (when provided) may be at the utility connection to your facility, at the substation, or a representative nearby site. But you may have sensitive equipment connected at a location inside your plant, where the sags may not be identical to those at the utility feed (due to the facility's electrical design). Your best approach is to measure sag data as close to the actual feed of the proposed equipment as possible. If you don't have actual location data, you'll have to use your best judgment for the data you can obtain.

If measured data isn't available, you can install monitors and collect the necessary sag data. In such a situation, the longer you collect it, the more reliable the data will be. You can find information on recommended methods and suggested equipment to collect power quality data in IEEE1159-1995, Recommended Practice for Monitoring Electric Power Quality.

The biggest influence on the frequency of sags is weather. A year with above average storms can significantly skew the data. Therefore, it may take a number of years of accumulated data to obtain an accurate picture of sag conditions at your site.

Predictive techniques use historical power failure rates to estimate the supply sag characteristics. IEEE 493-1997, Recommended Practice for the Design of Reliable Industrial and Commercial Power Systems (The Gold Book), provides guidance on predictive techniques for voltage sags. Such techniques allow you to analyze a variety of scenarios, while providing a means to estimate performance at new sites where there's no historical sag data. Predictive techniques may be helpful. But in almost all cases, utility data (if available) is more likely to be correct.

Without measured data and where lack of time or information precludes predictive analysis, you can use example data (someone else's utility sag data, national data, or other site data). Yes, such data doesn't establish a particular utility's service reliability. However, it does represent a sample of power system performance data.

Understanding the standard's methodology. IEEE1346-1998 does not intend to set performance criteria for utility systems, power distribution systems, or electronic process equipment. Instead, it shows how you can use the performance data for each of these to evaluate their compatibility as a system, in financial terms. The recommended methodology also provides standardization of methods, use of data, and analysis of power systems and equipment in evaluating compatibility within a common frame of reference.

You should apply the methodology at the planning or design stage of a project, when power supply and equipment choices are still flexible. The cost of trying to fix an incompatible system after equipment installation is more expensive than addressing the issue during the planning stage of a project, before specifying equipment. As such, this document does not discuss troubleshooting or correcting existing power quality problems.

Since voltage sags present the greatest financial loss due to compatibility, this first edition of the standard develops a compatibility methodology specifically for these events. However, compatibility encompasses many other issues, such as harmonics, surges, radiated interference, etc. As better information becomes available on power environment/equipment sensitivity response (and we gain experience with this approach), IEEE will develop compatibility methodologies for other issues.

This standard does not discuss technical options to improve compatibility. References and bibliographic listings in the standard provide detailed discussions of alternative methods. The corrective alternatives are so numerous and the sensitive equipment impacted by voltage sags so specific, such a listing would detract from the basic purpose of the document, which is to plan for compatibility.

Importance of equipment sensitivity to power anomalies. The new standard provides some information on the reaction of sensitive industrial and commercial equipment to voltage sags.

To cover a wide variety of equipment, it makes some broad generalizations. Therefore, we recommend you perform a detailed study of possible problems caused by installing sensitive electrical apparatus in a facility's electrical system during the initial design stage. Application needs and financial trade-offs will determine final equipment specifications and installation practices.

Your best source for information on voltage sag impact on electrical/electronic apparatus is the equipment manufacturer. Additionally, some independent sources, such as the Electric Power Research Institute (EPRI), test and collect data on various devices under voltage sag conditions. (See the new IEEE Standard's appendix.)

Test your sensitive equipment. These tests should relate to susceptibility of malfunction from the effect of single-, 2-, and 3-phase sags, phase shifting, and point-on-wave transition angles.

Since most electronic equipment achieves its voltage sag performance by capacitive energy (based on the peak sine wave value), the effect can be multiplied. For instance, a peak voltage depression of 10% will result in a 20% loss in energy storage capability, thus a lessening of voltage sag performan ce of the affected equipment.

Acknowledgment. The authors wish to recognize contributions made by members of IEEE Working Group 1346 toward the development of this standard. Special credit goes to Van E. Wagner, Staff Power Quality Engineer, Square D Co., who served as Chairman of the Working Group; and Larry E. Conrad, Manager of Operations Engineering/Power Quality, Cinergy Corp., who served as Chairman of Chapter 9 of IEEE 493-1997 (The Gold Book).




Sidebar: Compatibility Guidelines

When reviewing equipment specifications with a manufacturer, you should ask the following questions. Appropriate or inadequate answers usually indicate how knowledgeable the person is on the effect of compatibility.

How much can the supply voltage sag (and for how long) before equipment malfunctions?

Inadequate: A fixed value of some magnitude (e.g. 20% to 30%) with no mention of duration, phase shift, or unbalanced sag.

Preferred: A valid response would be in the form of a graph showing acceptable sag tolerances via magnitude (from nominal to 0) and duration (from 0 sec to 5 sec), as well as sensitivity to phase shift, voltage imbalance, and point on wave transition angles. On-site testing relating to sags (as a part of the acceptance testing of the installed complete process/machine) should be part of the purchase contract.

What are the steady-state voltage tolerances?

Inadequate: Plus or minus 5%. Typically, tight tolerances of this nature indicate someone is trying to improve ride-through characteristics using unrealistic criteria or covering for equipment containing parts with different nominal voltages.

Preferred: Values should be a minimum of +/-10% of the equipment rating. You must base all machine components on the same nominal voltage.

A 100V Asian-made computer with a 120V U.S. contactor having a 200V European drive will work; but only over a narrow-voltage bandwidth. How long can the composite machine tolerate zero voltage?

Inadequate: A time value without load or process dependency.

Preferred: The manufacturer should provide test results having graphs of the composite machine's operation. On-site tests covering this criterion (as a part of the acceptance testing of the installed complete machine) should be part of the purchase contract.

What is the machine's transient voltage withstand capability?

Inadequate: MOV arresters with "____" joules of protection protect this machine.

Preferred: The equipment has a fully coordinated transient suppression scheme, as demonstrated by curves of let-through voltage per amp of transient energy versus the withstand curves of electronic components and the BIL (Basic Impulse Level) of the internal circuitry. This scheme is adequate in an industrial environment with grounding sufficient for a frequency range of 3kHz to 1gHz.

How much voltage distortion can the equipment tolerate?

Inadequate: The equipment meets all standards, such as IEEE 519 and IEC 555.

Preferred: Voltage distortion tolerance of ,__%, crest factor of __, and notching of ,__volt-microseconds can be tolerated without affecting the equipment over a source impedance range of ____ohms.




Sidebar: Terms to Know

Notching: A switching (or other) disturbance of the normal power voltage waveform, lasting less than 0.5 cycles, which is initially of opposite polarity than the waveform and is thus subtracted from the normal waveform in terms of the peak value of the disturbance voltage.

PU: You use a per unit (pu) system where voltages and currents are expressed in percent or per unit of rated values. The corresponding base for expressing per unit reactances is the ohm producing a voltage drop of rated volts per phase, when rated current flows through the circuit. This procedure allows ease of comparison between electrical circuits or machines of different values. Time constants are expressed in seconds.

Phase shifting: The displacement in time of one waveform relative to another of the same frequency and harmonic content.

Point on wave transition angles: The points on a sine wave where the actual waveform deviates from the ideal waveform and returns to the ideal waveform. This has great impact on dropout magnitudes for magnetic contactors, relays, solenoids, and SCR-controlled equipment.

Sag (dip): A decrease in voltage or current at the power frequency for durations of 0.1 cycle to 1 min. Typical values are 0.1 to 0.9 pu. Note: To give a numerical value to a sag, the recommended standard usage is "a sag to 20%," which means the line voltage is reduced by 80%, to 20% remaining voltage (e.g. from 480V nominal, the voltage will be reduced to 96V). Using the preposition "of" (as in "a sag of 20%" or "a 20% sag") is discouraged.

Transient: Pertaining to or designating a phenomenon or a quantity that varies between two consecutive steady states during a short time interval compared with the time scale of interest. A transient can be a unidirectional impulse of either polarity or a damped oscillatory wave with the first peak occurring in either polarity.

Voltage magnitude imbalance: (As Applied To Sags Relating To Voltage Magnitude In Unbalance Polyphase Systems.) This is the absolute voltage difference between the highest and lowest phase voltages. This greatly affects 3-phase diode bridges and thus variable speed drives, including power supplies during single-phase sags. This differs from motor unbalance, which is maximum deviation from average.